We have focused our studies in a number of different areas: 1. Treg control the size of the memory-phenotype T cell population. During the course of studies in the LCMV model, we attempted to explore the role of Treg by specifically depleting Treg with Diptheria Toxin (DT) using mice that express the DT receptor exclusively on Foxp3+ Treg (Foxp3-DTR mice). Administration of DT to these mice results in the transient depletion of > 95% of Foxp3+ Treg. Although these studies were not useful in the LCMV experiments, the study of these mice affords us with an opportunity to examine the effects of polyclonal Treg in the steady state and to address some of the basic functions of Treg in the absence of inflammation or autoimmune disease. The Treg population was reconstituted within 72h following almost complete depletion when Foxp3-DTR mice were treated with DT. Treg reconstitution was mediated in part by IL-2, but the cellular source of the IL-2 remains to be determined. DT-induced Treg depletion in adult mice rapidly results in a three-fold expansion of activated memory phenotype CD44hiCD4+ T cells. Expansion was completely unaffected by the administration of anti-MHC Class II. The increase in proliferative rate was observed in a model where true memory T cells could be assessed by MHC Class II tetramer staining following acute LCMV infection. Depletion of Tregs also resulted in the proliferation of nave T cells presumably in response to autoantigens, as proliferation was completely blocked by anti-MHC class II. 2. For several years we have been interested in the role of cell surface and secreted TGF-beta in Treg Function. TGF-beta can have pleotropic effects on different cell types ranging from immune suppression, the promotion of fibrosis, to promotion/suppression of tumor growth. Activated regulatory T-cells (Tregs), but not activated T conventional cells, express the leucine rich repeat protein, GARP, which is responsible for surface localization of latent TGF-beta1. Although TGF-beta1 has been implicated in the suppressor function of Tregs, Treg conditional knock outs of TGF-beta1 or GARP display normal suppressor function in vitro. GARP-deficient Tregs develop normally, are present in normal numbers in peripheral tissues, but are unable to produce biologically active TGF-beta1, as measured by their inability to drive Th17 differentiation in vitro in the presence of IL-6 or to induce Foxp3+ Treg in the presence of IL-2. Both of these activities require the Treg-specific expression of integrin alphaVbeta8 that processes latent-TGF-beta1 to its activated form. However, the physiologic role of the GARP/latent TGF-beta1 on the surface of Tregs in vivo has remained elusive. Both GARP-deficient and TGF-beta1 deficient Tregs have a more activated phenotype in vivo suggesting that activation of latent TGF-beta1 by alphaVbeta8 functions in a cell intrinsic manner to modulate Treg function. Furthermore, conditional deletion of GARP in Treg results in significant changes to Treg subpopulations as measured by expression of Helios and Neuropilin-1, particularly in the gut-associated lymphoid tissues and mesenteric lymph nodes. Taken together, it appears that the GARP/latent TGF-beta1 complex plays an important role in Treg homeostasis in general and may modulate the differentiation and expansion of peripherally induced Treg. 3. The glucocorticoid-induced tumor necrosis factor related receptor (GITR), a member of the TNF receptor superfamily (TNFRSF) is expressed at high levels on the majority of freshly explanted Foxp3+ Treg cells, activated CD4+ and CD8+ T effector (Teff) cells and at low levels on other cell types including B cells, NK cells, macrophages, dendritic cells, eosinophils, basophils, and mast cells. The GITR ligand (GITR-L) is also widely expressed in the immune system and can be detected on basal levels on dendritic cells, B cells, monocytes, macrophages, with particularly high expression on plasmacytoid DCs. The expression of the GITR-L is transiently upregulated during inflammatory responses. Experiments using anti-GITR agonistic antibodies initially suggested that GITR played a critical role in the function of Treg cells, as engagement of the GITR by the agonist antibody appeared to reverse the suppressive effects of Treg cells in vitro. Subsequent studies using combinations of GITR-sufficient and deficient (KO) Treg cells and Teff cells in vitro demonstrated that the abrogation of suppression was secondary to engagement of the GITR on Teff cells rather than Treg cells, thereby rendering the Teff cells resistant to suppression. We have used a non-depleting, recombinant Fc-GITR-L and combinations of GITR WT and GITR KO Treg cells and Teff cells to re-examine the effects of GITR stimulation on each subpopulation in both un-manipulated mice and in a well-characterized model of inflammatory bowel disease (IBD). Treatment of mice with Fc-GITR-L resulted in significant expansion of Treg cells and a modest expansion of Tconv cells. When RAG KO mice were reconstituted with Tconv cells alone, GITR-L resulted in Tconv-cell expansion and severe IBD. The protective effect of Treg cells was lost in the presence of Fc-GITR-L, secondary to death of the Treg cells. When RAG KO mice were reconstituted with Treg cells alone, the transferred cells expanded normally, and Fc-GITR-L treatment resulted in a loss of Foxp3 expression, but the ex-Treg cells did not cause any pathology. Taken together, GITR activation on Treg cells can have different outcomes depending on the experimental context ranging from expansion in normal mice to death in the IBD model. This dual action of GITR engagement on Treg cells is not unexpected, as similar to other members of the TNFRSF, GITR might activate more than one signaling pathway. It also remains possible that the rapid induction of Treg-cell proliferation in a highly pro-inflammatory environment may result in activation induced cell death via FAS/FAS-L or TNF/TNFR. Taken together, the translation of studies of GITR function in the mouse model to the use of Fc-GITR-L or agonist mAbs in man should be undertaken with caution depending on the disease (autoimmunity vs. tumor immunity) under study and the immune status of the host.